Effect of Additional Treatments Combined with Conventional Therapies in Pregnant Patients with High-Risk Antiphospholipid Syndrome: A Multicentre Study

(1)

Effect of Additional Treatments Combined with

Conventional Therapies in Pregnant Patients

with High-Risk Antiphospholipid Syndrome:

A Multicentre Study

Amelia Ruffatti

1

Marta Tonello

1

Ariela Hoxha

1

Savino Sciascia

2

Maria J. Cuadrado

2

José O. Latino

3

Sebastian Udry

3

_{Tatiana Reshetnyak}

4

_{Nathalie Costedoat-Chalumeau}

5,6,7

_{Nathalie Morel}

5,6

Luca Marozio

8

Angela Tincani

9

Laura Andreoli

9

Ewa Haladyj

10

Pier L. Meroni

11,12,13

Maria Gerosa

11,12,13

_{Jaume Alijotas-Reig}

14,15

_{Sara Tenti}

16

_{Karoline Mayer-Pickel}

17

Michal J. Simchen

18

Maria T. Bertero

19

Sara De Carolis

20

Véronique Ramoni

21

Arsène Mekinian

22

Elvira Grandone

23

_{Aldo Maina}

24

_{Fátima Serrano}

25

_{Vittorio Pengo}

1,26

_{Munther A. Khamashta}

2,27,28

1_{Rheumatology Unit, Department of Medicine, University of Padua,} Padua, Italy

2_{Louise Coote Lupus Unit, Guy}_{’s and St Thomas’ NHS Foundation} Trust, St Thomas’ Hospital, London, United Kingdom

3_{Autoimmune,Thrombophilic Diseases and Pregnancy Division,} Dr. Carlos G Durand Hospital, Buenos Aires, Argentina

4_{Department of Systemic Rheumatic Disease, V.A. Nasonova} Research Institute of Rheumatology, Moscow, Russia 5_{AP-HP, Cochin Hospital, Internal Medicine Department, Centre de}

Référence Maladies Auto-immunes et Systémiques Rares, Paris, France 6_{Université Paris Descartes, Sorbonne Paris Cité, Paris, France} 7_{INSERM U 1153, Centre for Epidemiology and Statistics, Sorbonne}

Paris Cité (CRESS), Paris, France

8_{Department of Obstetrics and Gynaecology, Università di Torino,} Torino, Italy

9_{Rheumatology and Clinical Immunology Unit, Department of} Clinical and Experimental Sciences, University of Brescia, Spedali Civili, Brescia, Italy

10_{Department of Connective Tissue Diseases, National Institute of} Geriatrics, Rheumatology and Rehabilitation, Warsaw, Poland 11_{Division of Rheumatology, G. Pini Institute, Milan}

12_{Department of Clinical Sciences and Community Health, University} of Milan, Milan, Italy

13_{Istituto Auxologico IRCCS, Milan, Italy}

14_{Systemic Autoimmune Diseases Unit, Department of Medicine, Vall} d’Hebron University Hospital, Barcelona, Spain

15_{Department of Medicine, Faculty of Medicine, Universitat} Autònoma, Barcelona, Spain

16_{Department of Medicine, Surgery and Neuroscience,} Rheumatology Unit, University of Siena, Siena, Italy

17_{Department of Obstetrics, Medical University Graz, Graz, Austria} 18_{Maternal-Fetal Medicine, Department of Obstetrics and}

Gynecology, Sheba Medical Centre, Tel Hashomer, Israel 19_{Clinical Immunology, AO Ordine Mauriziano, Torino, Italy} 20_{Department of Obstetrics and Gynecology, Catholic University of}

Sacred Heart, Rome, Italy

21_{Division of Rheumatology, IRCCS Policlinico S. Matteo, Pavia, Italy} 22_{Service de Médecine Interne et l’inﬂammation-(DHU i2B), Hôpital}

Saint-Antoine, Université Paris 06, Paris, France

23_{Atherosclerosis and Thrombosis Unit, I.R.C.C.S. Casa Sollievo della} Sofferenza, S. Giovanni Rotondo, Foggia, Italy

24_{Servizio di Medicina Interna, Ospedale Sant’Anna, Torino, Italy} 25_{Department of Obstetrics and Gynecology, Centro Hospitalar de}

Lisboa Central, NOVA Medical School, Universidade Nova de Lisboa, Lisbon, Portugal

26_{Cardiology Clinic, Department of Cardiac, Thoracic and Vascular} Sciences, University of Padua, Padua, Italy

27_{Lupus Research Unit, The Rayne Institute, Division of Women}_’s Health, St Thomas Hospital, London, United Kingdom 28_{Department of Rheumatology, Dubai Hospital, Dubai,}

United Arab Emirates

Thromb Haemost 2018;118:639–646.

Address for correspondence Prof. Amelia Ruffatti, Reumatologia,

Palasanità, Via Giustiniani, 2 - 35128 Padova, Italy (e-mail: amelia.ruffatti@unipd.it).

received August 23, 2017 accepted after revision January 5, 2018

(2)

Introduction

A heparin–aspirin combination constitutes the conventional treatment protocol for pregnant women affected with anti-phospholipid syndrome (APS). As these strategies fail in approximately 20 to 30% of cases,1uncovering other options for women refractory to conventional treatment or at high risk of pregnancy complications has become an urgent under-taking. High-risk antiphospholipid antibody (aPL) profiles seem to be linked to specific serological markers such as multiple aPL positivity2–4and, in particular, to contempora-neous positivity to all three aPL assays,5–8 to lupus antic-oagulant (LAC) activity9or to high aPL titers.10,11Moreover, some well-defined clinical features such as a history of thromboembolism3,5,7,12–14and/or the presence of a systemic autoimmune disease7,9,15,16have been found to be associated with severe maternal–foetal complications in pregnant APS women receiving conventional therapy. Several experts are convinced that in association with conventional therapy, these high-risk APS patients should also be prescribed additional treatments before/during pregnancy in the effort to improve live birth rates and/or reduce pregnancy complications which often occur despite conventional treatment.14 Treatments prescribed in addition to conventional therapy, which cur-rently include intravenous immunoglobulins (IVIGs), low-dose steroids, plasma exchange or hydroxychloroquine (HCQ), have produced variable results.17,18

A retrospective, multicentre study recently reported that pregnant APS patients with previous thrombosis and triple aPL positivity treated with additional therapy had signiﬁcantly

higher live birth rates with respect to those receiving conven-tional therapy alone.18A variety of additional therapies includ-ing IVIG infusions, plasma exchange and low-dose steroids, alone or combined, were evaluated in that study, but it was impossible to analyse each therapy singularly, as the number of patients studied was insufficient to draw any significant conclusions.18The current, large, multicentre, observational, retrospective study set out in that study’s footsteps to inves-tigate the effect of various additional treatments on pregnancy outcomes in primary APS (PAPS) women refractory to con-ventional therapy and/or with risk factors for pregnancy complications to identify the most efficacious ones.

Materials and Methods

Study Population

Only patients who fulﬁlled—at the time they were diagnosed with PAPS—the clinical and laboratory classiﬁcation criteria established by the International Consensus in 200419were retrospectively enrolled in the study.

Patients fulﬁlling both the laboratory and clinical criteria were enrolled: (1) Laboratory criteria referred to laboratory risk factors, that is a positivity to LAC alone or associated with IgG/IgM anticardiolipin (aCL) and/or with IgG/IgM anti-β2-glycoprotein I (anti-β2GPI) antibodies. In accordance with data in the literature, the following types of aPL proﬁles were considered laboratory risk factors: triple aPL positivity5–8 and double/single aPL positivity always including the pre-sence of LAC.9At least two consecutive positive antibody results more than 12 weeks apart were needed to meet this

Keywords

► antiphospholipid

syndrome

► hydroxychloroquine

► low-dose steroids

► intravenous

immunoglobulins

► plasma exchange

Abstract

The effect of additional treatments combined with conventional therapy on pregnancy

outcomes was examined in high-risk primary antiphospholipid syndrome (PAPS)

patients to identify the most effective treatment strategy. The study

’s inclusion criteria

were (1) positivity to lupus anticoagulant alone or associated with anticardiolipin and/

or anti-

β2 glycoprotein I antibodies; (2) a history of severe maternal–foetal

complica-tions (Group I) or a history of one or more pregnancies refractory to conventional

therapy leading to unexplained foetal deaths not associated with severe maternal

–

foetal complications (Group II). Two different additional treatments were considered:

oral—low-dose steroids (10–20 mg prednisone daily) and/or 200 to 400 mg daily doses

of hydroxychloroquine and parenteral

—intravenous immunoglobulins at 2 g/kg per

month and/or plasma exchange. The study

’s primary outcomes were live birth rates

and pregnancy complications. A total of 194 pregnant PAPS patients attending 20

tertiary centres were retrospectively enrolled. Hydroxychloroquine was found to be

linked to a signi

ﬁcantly higher live birth rate with respect to the other oral treatments in

the Group II patients. The high (400 mg) versus low (200 mg) doses of

hydroxychlor-oquine (

p ¼ 0.036) and its administration before versus during pregnancy (p ¼ 0.021)

were associated with a signiﬁcantly higher live birth rate. Hydroxychloroquine therapy

appeared particularly ef

ﬁcacious in the PAPS patients without previous thrombosis.

Parenteral treatments were associated with a signi

ﬁcantly higher live birth rate with

respect to the oral ones (

p ¼ 0.037), particularly in the Group I patients. In conclusion,

some additional treatments were found to be safe and ef

ﬁcacious in high-risk PAPS

pregnant women.

(3)

requirement. (2) Clinical criteria referred to one or more of the following: a history of maternal thrombosis and/or previous severe pregnancy complications including eclamp-sia, severe preeclampsia (arterial pressure 160/110 and proteinuria 5 g in a 24-hour urine sample), haemolysis, elevated liver enzymes, low platelets (HELLP) syndrome, intrauterine growth restriction (IUGR—a postnatal birth weight less than the 10th percentile for the gestational age) and/or a previous pregnancy refractory to conventional therapy leading to an unexplained foetal death at or beyond the 10th week of gestation not associated with severe pregnancy complications.

Patients with any of the following criteria were excluded: (1) a previous conventionally treated pregnancy that led to a live birth; (2) the presence of a well-deﬁned associated autoimmune systemic disease including systemic lupus erythematosus; (3) age 45 years; (4) single or multiple organ failure; (5) severe pulmonary hypertension; (6) other known causes of pregnancy failure.

In addition to conventional therapy (prophylactic or therapeutic doses of heparinþ low-dose aspirin), the fol-lowing treatments were administered: oral treatments such as low-dose steroids (10–20 mg prednisone daily) and/or a 200 to 400 mg dose of HCQ daily or parenteral treatments such as IVIG (2 g/kg per month) and/or plasma exchange administered following a defined timetable. As no controlled clinical trials have as yet confirmed the efficacy of any additional therapy, treatment decisions tended to be based on the opinion and personal experience of the attending physician, and the patient’s clinical history in this context could seem more important than a particular aPL profile.

To facilitate statistical analysis, the patients studied were arbitrarily classiﬁed into two groups depending exclusively on their clinical features: 154 patients (79.4%) who pre-sented more severe clinical features such as a history of thrombosis and/or severe pregnancy complications includ-ing eclampsia, severe preeclampsia, HELLP syndrome and IUGR were assigned to Group 1; 40 patients (20.6%) who presented less severe clinical features such as a history of one or more pregnancies refractory to conventional treatment leading to unexplained foetal deaths not associated with severe pregnancy complications were assigned to Group II.

The study’s primary outcomes were the live birth rate (i.e., the number of live newborns surviving theﬁrst 27 days after birth) and severe pregnancy complications. The mean week of gestation at delivery, the mean birth weight in percentiles and the number of neonatal complications were the secondary outcomes.

The Institutional Review Board for Observational Studies and the Audit Committee of the University of Padua’s Medical Centre approved the study design. Patients who fulﬁlled the inclusion requirements were contacted and asked to sign informed consent forms. Their medical records were then retrieved and reviewed.

Antibody Detection

Five of the participating centres used a home-made enzyme-linked immunosorbent assay (ELISA) in accordance with the

minimal requirements proposed by the European Forum on Antiphospholipid Antibodies to determine aCL and anti-β2GPI antibodies of IgG and IgM isotypes.20,21The other 15 centres that determined IgG/IgM aCL and anti-β2GPI anti-bodies using commercial kits were recommended to follow the manufacturers’ directions in particular with regard to the less than 10% inter- and intra-run coefﬁcient of variability. LAC was assessed by multiple coagulation tests using plate-let-poor plasma samples, following internationally accepted guidelines.22,23

Statistical Analysis

The effects of each additional treatment were analysed sepa-rately in the two groups. The associations between the treat-ments and the primary outcomes were analysed using Pearson’s chi square test and Fisher’s exact test. The Krus-kal–Wallis test with multiple pairwise comparisons was used to analyse the continuous secondary outcomes. The Bonferroni correction was used in all the multiple comparisons in which the contingency table was larger than 2 rows 2 columns. Multiple comparisons of the primary and secondary outcomes in relation to the additional treatments were performed correcting thep-value using the Bonferroni adjustment. Back-ward conditional logistic regression analysis that was adjusted for confounding factors including disease duration, maternal age 35 years, congenital risk factors (factor V Leiden, pro-thrombin G20210A mutation, decrease in C and S protein antigens and activities, antithrombin III deﬁciency and increased activated protein C resistance), hypertension (blood pressure>140/90 mm Hg or use of antihypertensive drugs), body mass index30 kg/m2_{, aCL, anti-}_β

2GPI and other auto-antibodies was performed to evaluate the independent role of the additional treatments on the primary outcomes. Ap-value <0.05 was considered signiﬁcant. Statistical analysis was performed using the SPSS Statistics version 24 software.

Results

Between 1999 and 2016, a total of 194 pregnant PAPS patients (mean age¼ 32.05 years 5.02 SD, range ¼ 17– 44, mean disease duration¼ 5.01 years 4.5 SD, range ¼ 0–16) attending 20 international centres belonging to the European Forum on Antiphospholipid Antibodies net-work were retrospectively assessed. For 44 of the women (22.7%), it was theﬁrst pregnancy. One hundred ﬁfty (77.3%) had a history of pregnancy morbidity; in 81 cases (54%), the pregnancy ended in foetal death, 45 (30%) premature birth and 14 (9.3%) early miscarriage, and for 10 (6.6%) in a combination of these. The precedent pregnancy had been untreated in 18 cases (12%), and it was refractory to con-ventional therapy in 132 cases (88%).

One hundred twenty-seven of the PAPS patients (65.5%) had triple aPL positivity (IgG/IgM aCL plus IgG/IgM anti-β2GPI antibodies plus LAC) and 67 (34.5%) had LAC alone or associated with IgG/IgM aCL or IgG/IgM anti-β2GPI antibo-dies. Group I comprised 154 patients (79.4%): 76 (49.3%) had a history of thrombosis, 44 (28.6%) had previous severe pregnancy complications and 34 (22.1%) suffered from

(4)

both complications. Group II comprised 40 patients (20.6%): all had a previous foetal death refractory to conventional therapy not associated with severe obstetrical complications. All 194 pregnant PAPS patients were administered other treatments in addition to conventional therapy. One hundred forty-nine (76.8%) received oral additional treatments: 94 HCQ (63.1%), 36 low-dose steroids (24.2%) and 19 both (12.7%). Forty-ﬁve (23.2%) were prescribed parenteral treat-ments: 16 IVIG (35.5%), 8 plasma exchange (17.7%), and 21 both (46.6%). Additional treatments were initiated before pregnancy in 112 cases (57.7%) and during pregnancy in 82 (42.3%). Although no major side-effects were registered in association with the various types of treatments, one patient discontinued HCQ at the 12th week of gestation due to diffuse dermatitis.

One hundred sixty-three pregnancies (84%) ended favour-ably producing 164 live infants including one set of twins, all born between the 24 and 41 weeks of gestation (mean 35.7 3.3 SD). There were 115 (70.5%) caesarean deliveries and 48 (29.4%) vaginal ones. Thirty-ﬁve infants (21.3%) were born in less than 34 weeks of gestation and 91 (55.5%) infants were born in 37weeks of gestation. The infants (84 males and 80 females) had a mean birth weight of 2,638.2 g 733.9 SD (range: 600–3,950) and a mean of 42.2 percentiles 28.1 SD (range: 2–99). The infants’ mean Apgar score at 5 minutes was 8.8 1.2 SD (range: 3–10). Thirty-one (15.9%) of the pregnancies had negative outcomes: 19 foetal deaths (61.3%), 7 early miscarriages (22.6%) and 5 premature neonates (16.1%) who died during the perinatal period.

There were 59 cases (30.4%) of severe pregnancy compli-cations: 16 (27.1%) had eclampsia or severe preeclampsia, 16 (27.1%) had IUGR, 6 (10.2%) had HELLP syndrome, 3 (5.1%) had vascular thrombosis and 2 (3.4%) had catastrophic APS during puerperium. There were multiple complications in 16 cases (27.1%). There were 43 out of 149 (28.9%) pregnancy complications during oral treatment and 16 out of 45 (35.5%) during parenteral treatment; the difference was not signi ﬁ-cant (p ¼ 0.502). There was also no signiﬁcant difference in

the two patient groups in the frequency of each type of pregnancy complication.

Neonatal complications were recorded in 19 cases (11.6%): 12 had respiratory distress (63.2%), 4 had infections (21.0%), 1 had West’s syndrome (5.3%; in this case, the mother had received low-dose steroid), 1 had multiple malformations (5.3%; in this case, the mother had received low-dose steroid) and 1 had intra-lobar pulmonary seques-tration (5.3%), which was excised without complications 9 months after birth (in this case, the mother had been prescribed plasma exchange and IVIG).

The Effect of the Different Additional Treatments

The live birth rate and severe pregnancy complications (the primary outcomes) werefirst analysed in relation to the oral and parenteral treatments (►Table 1). Parenteral therapies produced a significantly higher live birth rate with respect to oral treatments (p ¼ 0.037), but only when the patients were considered together. The level of significance was higher in the group I patients (Group I:p ¼ 0.083 vs. Group II: p ¼ 0.596). As far as the primary outcomes linked to the different oral treatments (►Table 2) were concerned, HCQ therapy admi-nistered alone was associated with a significantly higher live birth rate with respect to both low-dose steroid alone and HCQþ low-dose steroid treatments in all the patients con-sidered together (p ¼ 0.027) as well as in the Group II patients considered separately (p ¼ 0.017). Severe preg-nancy complications were significantly less frequent in the HCQ-treated patients (p ¼ 0.008) with respect to the other oral therapies-treated patients. This latter HCQ’s effect was more evident in the Group II (p ¼ 0.008) than in the Group I (p ¼ 0.042) patients. Four hundred milligram daily doses of HCQ versus 200 mg doses (p ¼ 0.036) and its administration before versus during pregnancy (p ¼ 0.021) were associated with a significant increase in the live birth rate (►Table 3). When drug dosage was not considered, no statistically sig-nificant differences were found in the live birth rate between the patients who started HCQ before or during pregnancy

Table 1 Live birth rate and pregnancy complications in PAPS patients receiving additional oral and parenteral treatments Pregnant PAPS

patients

Oral treatments Parenteral treatments p

No. % No. %

Live birth rate

All patients: 194 121/149 81.2 42/45 93.3 0.037a Group I: 154 92/111 82.9 40/43 93.0 0.083 Group II: 40 29/38 76.3 2/2 100 0.596 Pregnancy complications All patients: 194 45/149 30.2 14/45 31.1 0.522 Group I: 154 37/111 33.3 12/43 27.9 0.327 Group II: 40 8/38 21.1 2/2 100 0.058

Abbreviation: PAPS, primary antiphospholipid syndrome.

Notes: Group I, PAPS patients with a history of thrombosis and/or severe pregnancy complications. Group II, PAPS patients with a history of refractory to conventional treatment foetal death occurring without any severe pregnancy complication.

a_{Statistical signi}_ﬁcance.

(5)

(p ¼ 0.462). It should be noted that HCQ taken before preg-nancy, which was administered at the time the pregnancy was being planned, led to a mean pregnancy duration of 37.1 weeks 2.4 SD; HCQ taken during pregnancy, which was administered at a mean gestational age of 9.5 weeks 7.3 SD, was linked to a mean pregnancy duration of 35.2 weeks 3.3 SD.

Unsuccessful pregnancies during HCQ treatment were sig-nificantly associated with previous thrombosis (p ¼ 0.025). In fact, 10 out of the 12 women (83.3%) receiving HCQ and suffering from a pregnancy failure had a history of thrombosis. Pregnancy losses in the other additional treatment patients were not, instead, significantly associated with a history of thrombosis. Six out of the 12 pregnancy losses (50%) occurring in the HCQ-treated women took place early in the pregnancy (before the 10th week of gestation), but only 1 out of the 19 (5.3%) pregnancy losses associated with other additional therapies occurred early. Spontaneous abortions were, thus, significantly more frequent in the women administered HCQ treatment with respect to the other additional treatment patients (p ¼ 0.0034). Five out of the six women (83.3%) suffering from early pregnancy loss during HCQ therapy were taking 200 mg daily doses.

When the primary outcomes in the different parenteral treatment groups were compared (►Table 4), no signiﬁcant

differences were found in the live birth rates. Severe pregnancy complications were signiﬁcantly less frequent in the patients treated with plasma exchange plus IVIG with respect to those treated with the other parenteral therapies both when the patients were considered together (p ¼ 0.034) as well as in the Group I (p ¼ 0.035) women analysed separately.

Theﬁndings regarding the secondary outcomes in the different additional therapies–treated patients are outlined in ►Table 5. According to the Kruskal–Wallis test, the patients receiving HCQ had signiﬁcantly later mean weeks of pregnancy, while those treated with plasma exchange alone had earlier ones (p < 0.001).

According to logistic regression analysis, IVIG adminis-tered alone (odds ratio [OR]¼ 0.123, 95% conﬁdence inter-val [CI]¼ 0.038–0.403, p ¼ 0.001), IVIG plus plasma exchange (OR¼ 0.135, 95% CI ¼ 0.033–0.550, p ¼ 0.005) and HCQ (OR¼ 0.450, 95% CI ¼ 0.207–0.978, p ¼ 0.044) alone had protective effects on severe pregnancy complica-tions in the Group I patients. Moreover, congenital risk factors (p ¼ 0.004) and body mass index 30 kg/m2 (p ¼ 0.019) were found to be independent risk factors for pregnancy complications in the Group II patients.

No statistical differences were found between the patients with triple aPL positivity and those with LAC alone or asso-ciated with IgG/IgM aCL or IgG/IgM anti-β2GPI antibodies. Table 2 Live birth rate and pregnancy complications linked to the three types of oral additional treatments

Pregnant PAPS patients

HCQ LDS HCQþ LDS p

No. % No. % No. %

Live birth rate

All patients: 149 82/94 87.2 27/36 75.0 12/19 63.1 0.027a Group I: 111 63/74 85.1 22/27 81.5 7/10 70.0 0.479 Group II: 38 19/20 95.0 5/9 55.5 5/9 55.5 0.017a Pregnancy complications All patients: 149 19/94 20.2 16/36 44.4 9/19 47.4 0.008a Group I: 111 19/74 25.7 14/27 51.8 4/10 40.0 0.042a Group II: 38 1/20 5 2/9 22.2 5/9 55.5 0.008a

Abbreviations: HCQ, hydroxychloroquine; HCQþ LDS, hydroxychloroquine plus low-dose steroid; LDS, low-dose steroid; PAPS, primary antipho-spholipid syndrome.

Table 3 Effect on live birth rate linked to two dosages of HCQ (200 or 400 mg) and the timing of its administration (before or during)

Live birth rate

HCQ dose Before pregnancy During pregnancy

400 mg 200 mg p 400 mg 200 mg p 400 mg 200 mg p

No. % No. % No. % No. % No. % No. %

47 94 35 79.5 0.036a ₃₄ _97.1 ₃₁ _79.5 _0.021a ₁₃ _86.7 ₃ ₇₅ _0.530

Abbreviation: HCQ, hydroxychloroquine.

a_{Statistical signiﬁcance.}

(6)

Discussion

This is thefirst study comparing the effect of different treat-ments administered in addition to conventional therapy to a large cohort of pregnant PAPS patients refractory to conven-tional therapy and/or at high risk of pregnancy complications. The patients studied were considered at high risk of preg-nancy-related complications on the basis of a clinical history of maternal–foetal complications and an aPL profile character-ized by the presence of LAC alone or associated with aCL and/or anti-β2GPI antibodies. The study’s most interesting findings concerned HCQ therapy which produced, mainly in the Group II patients, a significantly higher birth rate and significantly fewer severe pregnancy complications with respect to the other oral additional treatments. HCQ was found to be the most efficacious oral treatment used in association with conventional therapy in the PAPS patients with a previous foetal death refractory to conventional therapy.

Theseﬁndings are in agreement with those reported by two clinical studies examining the effect of HCQ in pregnant APS women that reported fewer pregnancy losses in the treated than in the untreated patients.24,25The novelty of the

current study regards the dosage and timing of HCQ treat-ment. In fact, the 400 mg dose of HCQ versus the 200 mg and its administration before but not during pregnancy were the two features that were associated with a significant increase in the live birth rate. HCQ therapy appeared particularly beneficial in the PAPS patients without previous thrombosis. In fact, the 10 out of the 12 unsuccessful pregnancies receiving HCQ treatment were significantly associated with a history of maternal thrombosis.

While it is true that only a few number of cases (45 cases) were treated with parenteral additional treatments, that therapy led to a signiﬁcantly higher live birth rate with respect to the oral medications, but only when all of the patients were considered together. As this result was mainly due to the contribution of the Group I patients, the more severe clinical subset, it suggests that these treatments should be reserved for PAPS patients with a history of severe maternal–foetal complications and/or in those refractory to HCQ additional treatment. In addition, logistic regression analysis showed that both IVIGþ plasma exchange and IVIG alone had a signiﬁcant protective effect against severe preg-nancy complications in the Group I patients.

Table 4 Live birth rate and pregnancy complications related to the three parenteral additional treatments Pregnant PAPS

patients

PE IVIG PEþ IVIG p

No. % No. % No. %

Live birth rate

All patients: 45 7/8 87.5 15/16 93.7 20/21 95.2 0.754

Group I: 43 7/8 87.5 13/14 92.8 20/21 95.2 0.765

Group II: 2 – 2/2 100 – n.a.

Pregnancy complications

All patients: 45 5/8 62.5 6/16 37.5 3/21 14.3 0.034a

Group I: 43 5/8 62.5 4/14 28.6 3/21 14.3 0.035a

Group II: 2 – 2/2 100 – n.a.

Abbreviations: IVIG, intravenous immunoglobulins; n.a., not applicable; PAPS, primary antiphospholipid syndrome; PE, plasma exchange; PEþ IVIG, plasma exchange plus intravenous immunoglobulins.

Table 5 Secondary outcomes related to the six additional treatments

Secondary outcomes HCQ LDS HCQþ LDS PE IVIG PEþ IVIG p

Week at delivery, mean SD 36.8 2.8 35.7 3 36.3 2.7 31.1 4.7 35.3 2.5 33.7 2.9 < 0.001 a Weight percentiles, mean SD 49.3 29.7 36.3 3 33.2 20.5 32.4 25.7 36.7 20.4 36.7 23.8 0.121 Neonatal complication, no. (%) 6 (7.3%) 5 (14%) 2 (16.7%) 3 (43%) 1 (6.7%) 2 (9.5%) 0.092

Abbreviations: HCQ, hydroxychloroquine; HCQþ LDS, hydroxychloroquine plus dose steroid; IVIG, intravenous immunoglobulins; LDS, low-dose steroid; PE, plasma exchange; PEþ IVIG, plasma exchange plus intravenous immunoglobulins.

a_{Statistical signiﬁcance.}

(7)

There were no major side effects correlated to the various types of treatments neither in the mothers nor in the children. As far as HCQ safety is concerned, our data are in accordance with those of a systematic review and a meta-analysis26 that recently reported that its use during preg-nancy was not associated with a relevant increase in con-genital malformations. In view of the fact that only a few studies have focused on HCQ treatment during PAPS preg-nancies, data from future investigations examining the long-term follow-up of mothers and children will provide more precise information about its safety in these patients. It is especially important to evaluate any side effects that could be linked to high dosage (400 mg daily) and/or long-term duration (before pregnancy).

While it is true that IVIG infusions and plasma exchange are indeed costly, the current study conﬁrmed that they are safe and lead to a high live birth rate in women who have little hope of having successful pregnancies.27–29 As this particular patient subset is quite rare, the expense of addi-tional parenteral treatments could be justiﬁed by the small number requiring such treatment.

One of the study’s limitations is its retrospective nature, but the rarity of these types of PAPS patients complicates performing prospective studies or clinical trials on addi-tional treatments. The fact that the study did not have a core reference laboratory could be considered another limitation of this study. LAC was nevertheless homoge-neously identiﬁed by each attending centre following international guidelines.22,23 As explained, homemade ELISA procedures and a variety of commercial kits were utilized by the various participating centres to determine aCL and anti-β2GPI antibodies. Despite many efforts to standardize the assays for these antibodies, methodologi-cal problems continue to persist and there are signiﬁcant inter-assay and interlaboratory variations in results even in connection to commercial kits.30–32It was thus impossible to compare all the methods used in the study in a standardized way to determine if they had any effect on the laboratory results. The strengths of the study could be its large number and the homogeneity of the study popu-lation composed only of PAPS patients refractory to con-ventional therapy and/or at high risk of pregnancy complications.

Our results will hopefully help point the direction of future clinical trials investigating the effect of additional treatments on refractory and high-risk PAPS patients with severe clinical history and LAC positivity. For the time being, the conclusion that can be drawn is that some additional treatments seem to be safe and efﬁcacious in refractory/ high-risk PAPS pregnant women. Four hundred milligram daily HCQ therapy begun before pregnancy may be a good option especially for PAPS patients with a previous preg-nancy refractory to conventional therapy leading to a foetal death not associated with severe pregnancy complications. IVIG combined with plasma exchange or alone could be used in very high-risk pregnant PAPS women and particularly in those with a history of thrombosis and/or refractory to previous HCQ additional treatment.

What is known about this topic?

• Since conventional therapeutic strategies fail in approximately 20 to 30% of pregnant patients affected with antiphospholipid syndrome (APS), there is an urgent need to deﬁne other therapeutic options to be used in addition to conventional treatment. • The most efﬁcacious additional treatment to be

com-bined with conventional therapy for high-risk APS pregnant women has not yet been established.

What does this paper add?

• Four hundred milligram daily hydroxychloroquine therapy begun before pregnancy may be a good option especially in APS patients with a previous, uncompli-cated, refractory to conventional therapy foetal death. • Intravenous immunoglobulins combined with plasma exchange or alone could be used in pregnant APS women with previous severe pregnancy complications.

Conﬂict of Interest

None.

Funding

None.

Acknowledgements

The authors would like to thank the following investiga-tors involved in the study: Brucato A., Cutolo M., Favaro M., Fioravanti A., Le Guern V., Meroni M., Montecucco C., Salvan E., Tabacco S., Villani M. and Inverso Moretti L. for editing the English version.

Finally, the authors would like to thank Dr. Branch D.W. for his critical review of the manuscript.

References

1 Lassere M, Empson M. Treatment of antiphospholipid syndrome in pregnancy–a systematic review of randomized therapeutic trials. Thromb Res 2004;114(5-6):419–426

2 Sailer T, Zoghlami C, Kurz C, et al. Anti-beta2-glycoprotein I antibodies are associated with pregnancy loss in women with the lupus anticoagulant. Thromb Haemost 2006;95(05):796–801 3 Fischer-Betz R, Specker C, Brinks R, Schneider M. Pregnancy outcome in patients with antiphospholipid syndrome after cere-bral ischaemic events: an observational study. Lupus 2012;21 (11):1183–1189

4 Matsuki Y, Atsumi T, Yamaguchi K, et al. Clinical features and pregnancy outcome in antiphospholipid syndrome patients with history of severe pregnancy complications. Mod Rheumatol 2015; 25(02):215–218

5 Ruffatti A, Tonello M, Del Ross T, et al. Antibody proﬁle and clinical course in primary antiphospholipid syndrome with pregnancy morbidity. Thromb Haemost 2006;96(03):337–341

6 Ruffatti A, Tonello M, Cavazzana A, Bagatella P, Pengo V. Labora-tory classiﬁcation categories and pregnancy outcome in patients with primary antiphospholipid syndrome prescribed antithrom-botic therapy. Thromb Res 2009;123(03):482–487

(8)

7 Ruffatti A, Tonello M, Visentin MS, et al. Risk factors for pregnancy failure in patients with anti-phospholipid syndrome treated with conventional therapies: a multicentre, case-control study. Rheu-matology (Oxford) 2011;50(09):1684–1689

8 Latino JO, Udry S, Aranda FM, Perés Wingeyer SDA, Fernández Romero DS, de Larrañaga GF. Pregnancy failure in patients with obstetric antiphospholipid syndrome with conventional treat-ment: the inﬂuence of a triple positive antibody proﬁle. Lupus 2017;26(09):983–988

9 Lockshin MD, Kim M, Laskin CA, et al. Prediction of adverse pregnancy outcome by the presence of lupus anticoagulant, but not anticardiolipin antibody, in patients with antiphospholipid antibodies. Arthritis Rheum 2012;64(07):2311–2318

10 Nodler J, Moolamalla SR, Ledger EM, Nuwayhid BS, Mulla ZD. Elevated antiphospholipid antibody titers and adverse pregnancy outcomes: analysis of a population-based hospital dataset. BMC Pregnancy Childbirth 2009;9:11

11 Simchen MJ, Dulitzki M, Rofe G, et al. High positive antibody titers and adverse pregnancy outcome in women with antiphospholipid syndrome. Acta Obstet Gynecol Scand 2011;90(12):1428–1433 12 Lima F, Khamashta MA, Buchanan NM, Kerslake S, Hunt BJ, Hughes

GR. A study of sixty pregnancies in patients with the antipho-spholipid syndrome. Clin Exp Rheumatol 1996;14(02):131–136 13 Bramham K, Hunt BJ, Germain S, et al. Pregnancy outcome in

different clinical phenotypes of antiphospholipid syndrome. Lupus 2010;19(01):58–64

14 de Jesús GR, Rodrigues G, de Jesús NR, Levy RA. Pregnancy morbidity in antiphospholipid syndrome: what is the impact of treatment? Curr Rheumatol Rep 2014;16(02):403

15 Bowman ZS, Wünsche V, Porter TF, Silver RM, Branch DW. Prevalence of antiphospholipid antibodies and risk of subsequent adverse obstetric outcomes in women with prior pregnancy loss. J Reprod Immunol 2015;107:59–63

16 Danowski A, de Azevedo MN, de Souza Papi JA, Petri M. Determi-nants of risk for venous and arterial thrombosis in primary antiphospholipid syndrome and in antiphospholipid syndrome with systemic lupus erythematosus. J Rheumatol 2009;36(06): 1195–1199

17 Ruffatti A, Hoxha A, Favaro M, et al. Additional treatments for high-risk obstetric antiphospholipid syndrome: a comprehensive review. Clin Rev Allergy Immunol 2017;53(01):28–39

18 Ruffatti A, Salvan E, Del Ross T, et al. Treatment strategies and pregnancy outcomes in antiphospholipid syndrome patients with thrombosis and triple antiphospholipid positivity. A European multicentre retrospective study. Thromb Haemost 2014;112(04): 727–735

19 Miyakis S, Lockshin MD, Atsumi T, et al. International consensus statement on an update of the classiﬁcation criteria for deﬁnite antiphospholipid syndrome (APS). J Thromb Haemost 2006;4 (02):295–306

20 Tincani A, Allegri F, Balestrieri G, et al. Minimal requirements for antiphospholipid antibodies ELISAs proposed by the Euro-pean Forum on antiphospholipid antibodies. Thromb Res 2004; 114(5-6):553–558

21 Reber G, Tincani A, Sanmarco M, de Moerloose P, Boffa MC; Standardization Group of the European Forum on Antiphospho-lipid Antibodies. Proposals for the measurement of anti-beta2-glycoprotein I antibodies. Standardization group of the European Forum on Antiphospholipid Antibodies. J Thromb Haemost 2004; 2(10):1860–1862

22 Brandt JT, Triplett DA, Alving B, Scharrer I. Criteria for the diagnosis of lupus anticoagulant. Thromb Haemost 1995;74 (04):1185–1190

23 Pengo V, Tripodi A, Reber G, et al; Subcommittee on Lupus Anticoagulant/Antiphospholipid Antibody of the Scientiﬁc and Standardisation Committee of the International Society on Thrombosis and Haemostasis. Update of the guidelines for lupus anticoagulant detection. J Thromb Haemost 2009;7(10): 1737–1740

24 Mekinian A, Lazzaroni MG, Kuzenko A, et al; SNFMI and the European Forum on Antiphospholipid Antibodies. The efﬁcacy of hydroxychloroquine for obstetrical outcome in anti-phospholipid syndrome: Data from a European multicenter retrospective study. Autoimmun Rev 2015;14(06):498–502

25 Sciascia S, Hunt BJ, Talavera-Garcia E, Lliso G, Khamashta MA, Cuadrado MJ. The impact of hydroxychloroquine treatment on pregnancy outcome in women with antiphospholipid antibodies. Am J Obstet Gynecol 2016;214(02):273.e1–273.e8

26 Kaplan YC, Ozsarfati J, Nickel C, Koren G. Reproductive outcomes following hydroxychloroquine use for autoimmune diseases: a systematic review and meta-analysis. Br J Clin Pharmacol 2016;81 (05):835–848

27 Cherin P, Marie I, Michallet M, et al. Management of adverse events in the treatment of patients with immunoglobulin ther-apy: a review of evidence. Autoimmun Rev 2016;15(01):71–81 28 El-Haieg DO, Zanati MF, El-Foual FM. Plasmapheresis and

preg-nancy outcome in patients with antiphospholipid syndrome. Int J Gynaecol Obstet 2007;99(03):236–241

29 Ruffatti A, Favaro M, Hoxha A, et al. Apheresis and intravenous immunoglobulins used in addition to conventional therapy to treat high-risk pregnant antiphospholipid antibody syndrome patients. A prospective study. J Reprod Immunol 2016;115:14–19 30 Reber G, Tincani A, Sanmarco M, de Moerloose P, Boffa MC; European Forum on Antiphospholipid Antibodies Standardiza-tion Group. Variability of anti-beta2 glycoprotein I antibodies measurement by commercial assays. Thromb Haemost 2005;94 (03):665–672

31 Pengo V, Biasiolo A, Bison E, Chantarangkul V, Tripodi A; Italian Federation of Anticoagulation Clinics (FCSA). Antiphospholipid antibody ELISAs: survey on the performance of clinical labora-tories assessed by using lyophilized afﬁnity-puriﬁed IgG with anticardiolipin and anti-beta2-Glycoprotein I activity. Thromb Res 2007;120(01):127–133

32 Favaloro EJ, Wheatland L, Jovanovich S, Roberts-Thomson P, Wong RC. Internal quality control and external quality assurance in testing for antiphospholipid antibodies: Part I–Anticardiolipin and anti-β2-glycoprotein I antibodies. Semin Thromb Hemost 2012;38(04):390–403